intellectual virtue perspectives from ethics and epistemology nov 2003

reliable belief-forming process² or faculty³ or agent.⁴ More robust forms ofvirtue epistemology make the fundamental bearer of epistemic value anepistemic or intellectual virtue in the s

Virtue ethics has attracted a lot of attention over the past few decades, andmore recently there has been considerable interest in virtue epistemology as

an alternative to traditional approaches in that field Ironically, althoughvirtue epistemology got its inspiration from virtue ethics, this is the first bookthat brings virtue epistemologists and virtue ethicists together to contributetheir particular expertise, and the first that is devoted to the topic of intellec-tual virtue

All new and right up to date, the papers collected here by Zagzebski andDePaul demonstrate the benefit of each branch of philosophy to the other

Intellectual Virtue will be required reading for anyone working in either field

Michael DePaul is Professor of Philosophy at the University of Notre Dame,

Indiana

Linda Zagzebski is Kingfisher College Chair of the Philosophy of Religion

and Ethics at the University of Oklahoma

Intellectual Virtue

Perspectives from Ethics and Epistemology

Edited by

Michael DePaul and Linda Zagzebski

CLARENDON PRESS OXFORD

Great Clarendon Street, Oxford ox2 6dp

Oxford University Press is a department of the University of Oxford.

It furthers the University’s objective of excellence in research, scholarship,

and education by publishing worldwide in

Oxford New York Auckland Bangkok Buenos Aires Cape Town Chennai Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi

São Paulo Shanghai Taipei Tokyo Toronto

Oxford is a registered trade mark of Oxford University Press

in the UK and in certain other countries

Published in the United States

by Oxford University Press Inc., New York

© the several contributors, 2003 The moral rights of the authors have been asserted

Database right Oxford University Press (maker)

First published 2003 First published in paperback 2007

All rights reserved No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press,

or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department,

Oxford University Press, at the address above

You must not circulate this book in any other binding or cover and you must impose this same condition on any acquirer British Library Cataloguing in Publication Data

Data available Library of Congress Cataloging in Publication Data

Data available ISBN 978-0-19-925273-2 (Hbk.) 978-0-19-921912-4 (Pbk.)

1 3 5 7 9 10 8 6 4 2

Printed in Great Britain

on acid-free paper by

T J International Ltd., Padstow, Cornwall

Typeset by Laserwords Private Limited, Chennai, India

3

List of Contributors vii

Linda Zagzebski and Michael DePaul

PART ONE CLASSICAL VIRTUE ETHICS AND

Julia Annas

2 Intellectual Virtue: Emotions, Luck, and the Ancients 34

Nancy Sherman and Heath White

PART TWO CONTEMPORARY VIRTUE ETHICS AND

John Greco

6 Intellectual Motivation and the Good of Truth 135

Linda Zagzebski

7 The Place of Truth in Epistemology 155

PART FIVE APPLYING VIRTUE TO EPISTEMOLOGY:

Robert C Roberts and W Jay Wood

Julia Annas, Regents Professor of Philosophy, University of Arizona

Michael DePaul, Professor of Philosophy, University of Notre Dame

J L A Garcia, Professor of Philosophy, Boston College

John Greco, Associate Professor of Philosophy, Fordham University

Christopher Hookway, Professor of Philosophy, University of Sheffield Christine McKinnon, Associate Professor of Philosophy, Trent University Wayne D Riggs, Associate Professor of Philosophy, University of Oklahoma Robert C Roberts, Distinguished Professor of Ethics, Baylor University Nancy Sherman, University Professor, Georgetown University

David Solomon, Associate Professor of Philosophy, University of Notre

Dame; Director of the Notre Dame Center for Ethics and Culture

Ernest Sosa, Romeo Elton Professor of Natural Theology and Professor of

Philosophy, Brown University; Distinguished Visiting Professor, RutgersUniversity

Heath White, Visiting Assistant Professor, Valparaiso University

W Jay Wood, Associate Professor of Philosophy, Wheaton College

Linda Zagzebski, Kingfisher College Chair of the Philosophy of Religion and

Ethics and Professor of Philosophy, University of Oklahoma

Linda Zagzebski and Michael DePaul

The concept of a virtue has been enormously important in ethics since its ning, but it has only recently been adopted by epistemologists In 1980 ErnestSosa introduced the idea of virtue into epistemological discourse in his paper

begin-‘The Raft and the Pyramid’.¹ Sosa’s motive for an interest in virtue arose out ofthe epistemological concerns of the time, in particular, the dispute betweenfoundationalists and coherentists, and it is quite different from the ethicist’smotive But Sosa’s idea signalled the beginning of a movement that came to becalled virtue epistemology At a minimum, virtue epistemology is characterized

by a shift in focus from properties of beliefs to the intellectual traits of agents.The primary bearer of epistemic value is a quality of the agent that enables her

to act in a cognitively effective and commendable way Some virtue ogists claim that traditional targets of epistemological investigation such asknowledge, rationality, or justification can be defined in terms of intellectualvirtue, whereas others argue that the traditional targets themselves ought to bereplaced by an investigation of virtue in the cognitive domain

epistemol-The earliest form of virtue epistemology was reliabilism According to theories of this kind, the basic component of knowledge or justified belief is a

¹ Sosa (1980).

reliable belief-forming process² or faculty³ or agent.⁴ More robust forms ofvirtue epistemology make the fundamental bearer of epistemic value anepistemic or intellectual virtue in the sense of virtue used in ethics,⁵ or theymay even model the structure of an epistemological theory on virtue ethics.⁶The alternatives for this last approach are as diverse as the varieties of virtue ethics and most of them are as yet unexplored There is also the altern-ative of eschewing theory altogether and adopting an anti-theory model forepistemology.⁷

Virtue epistemology is a recent movement, but virtue ethics is as old asWestern philosophy Ever since Plato, ethicists and historians of ethics haveexplored the nature of a virtue and the particular virtues, as well as the relationship between the concept of virtue and other key concepts in ethicssuch as that of a right act, a good motive, emotion, and happiness Virtue epistemologists understandably concentrate on the ways the idea of a virtuecan help resolve epistemological questions and leave the conceptual work ofexplaining value to ethics Clearly, then, virtue epistemology needs virtueethics But the editors of this volume believe that virtue ethics also has some-thing important to learn from virtue epistemology Perhaps due to historicalaccident, virtue ethicists have had little to say about intellectual virtue Theygenerally take for granted that the moral and intellectual virtues are not onlydistinct, but relatively independent Some may also think that it is the job ofsome other branch of philosophy to examine the intellectual virtues Granted,

Aristotle linked the moral virtues with the intellectual virtue of phronesis,

or practical wisdom, and for that reason Aristotle scholars and ethicists

influenced by Aristotle have attended to phronesis in their treatments of virtue,

but their interest is generally limited by their concern with the connection of

phronesis to the distinctively moral virtues They typically give no attention to

² Alvin Goldman has proposed a form of process reliabilism He does not use the term ‘virtue’

very often, but it appears in Goldman (1993a).

³ Ernest Sosa has proposed versions of faculty reliabilism in many places See Sosa (1991) for a collection of his papers Alvin Plantinga’s theory of warrant as proper function also appears in

many places, in particular, Plantinga (1993b) Plantinga is sometimes classified as a virtue

epistemo-logist, although he does not use the term ‘virtue’ for properly functioning faculties.

⁴ John Greco has recently proposed a theory he calls agent reliabilism in Greco (1999a) In that

paper Greco uses the term ‘agent reliabilism’ for a larger class of theories than his own, including Sosa’s, Plantinga’s, and Zagzebski’s early theory.

⁵ See Code (1987), Montmarquet (1993), and Zagzebski (1996) ⁶ See Zagzebski (1996).

⁷ The anti-theory movement has had a following among virtue ethicists For a collection of papers on this movement see Clarke and Simpson (1989).

such intellectual virtues as intellectual carefulness, thoroughness, humility,courage, trust, autonomy, or fairness As a matter of fact, virtue epistemologistshave not gone very far in investigating the individual intellectual virtues either,but they have taken the lead in addressing intellectual virtue as a topic of interest

and importance apart from the relationship between phronesis and moral virtue,

and some have begun a study of the relationship between the way in which weform beliefs and the way we conduct ourselves in our moral lives

We believe that the nature of intellectual virtue and vice is critical for thepurposes of both ethicists and epistemologists It is therefore ironic that there hasbeen so little interaction between them In an effort to remedy this problem,Michael DePaul organized a conference at the University of Notre Dame inSeptember 2000, which brought ethicists and epistemologists together to investi-gate the nature of intellectual virtue and its role in resolving disputes in ethicsand epistemology DePaul asked Linda Zagzebski to co-edit a book coming out ofthe conference, and additional papers by Wayne Riggs and Christine McKinnonwere added to the nine papers presented there This volume is the result of thatproject Some of these essays are written by philosophers whose work is prima-rily in ethics: Julia Annas, David Solomon, Jorge Garcia, and ChristineMcKinnon Some are written by philosophers whose work is primarily in episte-mology: Ernest Sosa, John Greco, Christopher Hookway, and Wayne Riggs Oneauthor (Linda Zagzebski) works in both epistemology and ethics The final twochapters are co-authored, with one author in ethics and the other in epistemol-ogy (Nancy Sherman and Heath White, and Robert C Roberts and W Jay Wood).The editors believe that intellectual virtue is one of the most promising topics in philosophy, but the literature on the topic is generally splintered intowork that is primarily concerned with historical scholarship, work intendedfor moral philosophers, and work intended for epistemologists As far as weknow, these are the first essays written by virtue epistemologists and virtueethicists in consultation with each other, including virtue ethicists with a historical orientation Epistemologists and ethicists bring different knowledgeand perspectives to the topic, and we think that the essays collected heredemonstrate the benefit of each branch of philosophy to the other

Traditional virtue ethics is usually associated with Plato and Aristotle, butStoic virtue ethics gets at least as much attention from the ethicists in this volume (Annas, and Sherman and White), resulting in a chapter adopting aStoic approach to virtue epistemology (Riggs) Two ethicists address contem-porary virtue ethics, either in its debate with consequentialism (Garcia) or inits attempt to be interestingly different from traditional approaches

(Solomon) Advanced discussions by ethicists on virtue ethics and its place inthe pantheon of ethical theories (or anti-theories) is important for virtue epistemologists who generally have not gone very far in investigating theplace of the different forms of virtue epistemology in the taxonomy of normative epistemological theories.

Moral philosophers have traditionally investigated the individual virtueswith great care, and in this volume Roberts and Wood’s fascinating chapter onintellectual humility gives a detailed investigation of this virtue for the purposes of epistemology We look forward to more inquiries of this kind.The epistemological chapters in the volume focus on a number of questions that expand the topics typically addressed by epistemologists Oneissue that gets considerable attention is the nature and scope of epistemicvalue Two chapters (Sosa and Zagzebski) address the problem of what makesknowledge more valuable than mere true belief, and one other discusses itbriefly (Greco) Some epistemologists are beginning to say that knowledge hasreceived too much attention in contemporary epistemology, and other epistemic values have been neglected In his chapter in this collection, Riggsargues that there is a need to expand the range of epistemic value to includeunderstanding and wisdom

The varieties of virtue epistemology and its potential for broadening thestandard set of problems in the field are addressed by Christopher Hookway.Some virtue epistemologists have previously argued that the concept of intel-lectual virtue can be used in solving such traditional epistemological problems

as the task of defining knowledge (Zagzebski, Greco, and Sosa) or answeringscepticism (Greco), whereas others claim that the real virtue of virtue epistemo-logy is the way it permits us to redefine the central questions The chapters

by Hookway and Riggs defend this position

Christine McKinnon argues for the advantages of applying feminist ethics

to epistemology since it permits an account of a broader range of cases ofknowing than those standardly discussed, in particular, knowledge of oneselfand others She argues that a virtue approach in epistemology is better suited

to giving an account of knowledge of persons than traditional approaches

1 Summaries of Essays

Julia Annas begins her chapter, ‘The Structure of Virtue’, by acknowledging the

interest of recent efforts to use a rich notion of virtue in epistemology She is

concerned, however, about reliance on Aristotle’s particular version of virtueethics to the exclusion of the rest of the ancient tradition She examines twoissues: the connection between virtue and skill and the relation of virtue tosuccess It turns out that the consensus position of ancient virtue ethics onthese issues differs from Aristotle’s in ways that are significant for the application

of the notion of virtue in epistemology Unlike Aristotle, the rest of theancient tradition held that moral virtue is a kind of skill, according to theStoics, the skill of living Moral virtue shares the same intellectual structure asother skills Intellectual virtues also share this structure, and hence are skills.But according to Annas the intellectual virtues are also importantly differentfrom the moral virtues While the moral virtues aim at doing the right thing,the intellectual virtues aim at truth These aims might converge, but theyneed not—indeed, they can conflict Hence, the intellectual virtues cannotsimply be subsumed under the moral virtues; the relations between them aremore complex Virtue clearly requires success, but the issue is complicated

since the virtuous person has two aims in acting The overall aim, or telos, is to

live a certain kind of life, one that is virtuous But each particular action also

has an immediate target, or skopos Which aim must be attained for a person to

have the kind of success necessary for virtue? Annas maintains Aristotle wasconfused here, but the Stoics were clear and answered that it is attainment ofthe ultimate aim Knowledge is different In order to know one must attain theimmediate aim of forming a true belief Hence, one cannot define knowledgesimply in terms of virtue

Nancy Sherman and Heath White point out that virtue epistemologists have

underutilized some of the key resources of classical virtue ethics, in particular,the role of affect in intellectual virtue, and the role of luck and external goods

in achieving knowledge Their chapter, ‘Intellectual Virtue: Emotions, Luck,and the Ancients’, begins by exploring the role of emotion in intellectualvirtue, and they defend the Aristotelian position that even though beliefs arenot fully voluntary because the emotions that influence them are not fullyvoluntary, they are within the reach of responsibility We are not primarilypassive with respect to our emotions Revising the cognitive core of emotions

is one of the ways we revise emotions themselves This is an Aristotelian point,but it is developed by the Stoics whose view of emotions was more thoroughlycognitive The Stoics viewed emotions as voluntary assents to appearances ofgood and evil They are judgements, but they are mistaken How, then, couldthe Stoics endorse emotion as a central aspect of cognitive character? Theanswer, say Sherman and White, is that the sage can resist being taken in by

appearances Further, there are affective states that dispose the agent to makeaccurate judgements To care about truth and certainty, in Stoic terms, is to benon-rash and non-careless in giving and withholding assent to the appearances.Since these are emotional attitudes, the affective component of intellectualvirtue found in Aristotle survives even the Stoic revision Aristotle and theStoics had contrasting positions on the place of luck in happiness, however,since the Stoics maintained and Aristotle denied that virtue is sufficient forhappiness The Stoics even tried to deny the place of luck in knowledge.Cognitive virtues are sufficient for getting the truth This position has beengiven up by modern epistemologists who almost always agree that there is asubstantial amount of luck in getting truth Sherman and White concludethat the difference between truth and happiness in the role of luck limits theextent to which virtue ethics and virtue epistemology can be unified.

In ‘Virtue Ethics: Radical or Routine?’ David Solomon sees the turn to virtue

ethics in the latter half of the twentieth century as taking two different forms.One focuses on the ordering of evaluative concepts and argues that the concept of virtue is more basic than the concepts of a right act and a good state

of affairs Solomon calls this routine because of its focus on familiar argumentsover theory construction The other form focuses on deeper questions aboutthe nature and ambition of modern ethics and its ability to satisfy our need forreflective guidance This more radical approach includes such themes as a suspicion of rules and principles, the importance of the narrative structure of

a human life, the importance of community, a critique of modernity, andsometimes a suspicion of moral theory itself Debates over virtue ethics so farseem unresolvable because they are partly debates over the criteria by which

an ethical theory should be judged Virtue epistemologists should be awarethat when they look to virtue ethics for a model, there are two very differentmodels to which they can appeal Solomon suggests that epistemologistsmight learn from the experience of moral philosophers about the variety ofuses to which the language of virtue can be put and possible confusions aboutthese uses

Jorge Garcia argues in ‘Practical Reason and its Virtues’ that the instrumentalist

conception of practical reasoning favoured by consequentialists is inadequateand incapable of protecting us against the moral horrors of the twentieth century This is even true of the sophisticated consequentialism of AmartyaSen, who proposes that human sympathy in combination with instrumentalreason is a safeguard against atrocities But this leaves us with the need to justifyacting from sympathy which, from the standpoint of instrumental reason,

may seem imprudent Sen maintains that the badness of rights-violations is anindependent badness which makes acts that produce it wrong if not counter-balanced by good outcomes Garcia argues that this still leaves us little protectionagainst gross injustice, which almost always is seen as arising from a kind ofsympathy—sympathy for humanity as a whole The problem with that,Garcia argues, is that genuine sympathy is always for individuals, and it is individuals who are the bearers of rights.

Garcia’s alternative is a theory of the moral life that has four characteristics:(1) It is role-centred, which means it makes all moral features (rights, virtues,duties) ones that a person has in virtue of being in role-relationships with others: friend, parent, fellow citizen, informant, and so on (2) It is virtues-based, which means that it makes judgements of right and wrong, rights andduties depend on more fundamental judgements of attitudinal responses thatare virtuous or vicious (3) It is patient-focused in that the fundamental attitudes of virtue are those directed towards the person with whom the agent

is related in the relevant role (4) It is input-driven, which is to say that themoral status of an act is determined by its motivational input, not the physicalstructure of the act or its consequential output Garcia argues that these features not only protect against tyranny but are sensitive to the moral signi-ficance of differentially demanding roles This is true of our epistemic roles aswell The intellectual virtues are neither instrumentally nor intrinsicallygood Like the moral virtues they are good-making in that they contributetowards our being good reasoners in the roles we have in our epistemic communities

Two kinds of problem have plagued fallibilists regarding knowledge: thelottery problem and Gettier problems In ‘Knowledge as Credit for True

Belief’, John Greco argues that we can resolve both kinds of problem by

attend-ing to the illucutionary force of knowledge attributions, specifically, that theyserve to give credit to the believer for getting things right The idea is that insaying someone knows we are saying that the person has formed a true belief

in virtue of her own effort and ability, and not because of some sort of goodfortune Greco begins his essay with sections devoted to each of the two kinds

of problem and failed efforts to address them He then takes up the task ofdeveloping his own account Using work done by Joel Feinberg on blaming,which stresses the assignment of causal responsibility, Greco develops a generalaccount of giving credit According to this account giving credit cruciallyinvolves assigning causal responsibility to the agent, not in the sense thatthe agent is picked out as the sole cause, but in the sense that the agent is

identified as a salient, or the most salient, part of the cause Since salience issensitive to context in various ways, Greco’s resolution of Gettier and lotteryproblems inherits a significant contextual element In addition to requiringthat the agent be causally responsible for something in order to get credit for

it, Greco requires that a relevant aspect of the agent’s character play a cant causal role A bumbling athlete who only rarely succeeds at some featwill not get credit even when she does, according to Greco, since the raresuccess will be attributed to good luck rather than the athlete’s skill What thiscomes to in the cognitive domain is that a believer’s reliable cognitive char-acter, or intellectual virtue, must be an important necessary element in thecause of a true belief for the believer to get credited with the true belief Afterpresenting his account, Greco tests it against a number of cases and closes with

signifi-a brief considersignifi-ation of how his signifi-account might help us understsignifi-and the vsignifi-alue

of knowledge

In previous work the editors of this volume have discussed the problem ofwhat makes knowledge better than true belief.⁸ Zagzebski calls this the value

problem In ‘Intellectual Motivation and the Good of Truth’, Linda Zagzebski

investigates the value problem further She distinguishes four ways a belief can

be evaluated according to its relation to truth: (1) A belief can have valuebecause truth is its consequence (2) A belief can have teleological value in theAristotelian sense, the kind of value something has when it is a component of

a good natural end On this account true belief would be intimately related to

the good of eudaimonia or a good life (3) A belief can be valuable in that truth is

its end in the sense of an aim Assuming that true belief is good, it is also good

to aim at it (4) A belief can be good because it arises from good motives, in particular, the motive of valuing truth or disvaluing falsehood Since motivesand aims are not the same thing, the fourth way in which the value of truth

is related to the value of a given true belief does not reduce to the third.Zagzebski argues that the fourth way in which a given belief can be related totruth makes the belief better than either the first or the third way She defendsthis claim by comparing beliefs to acts An act that aims at relieving suffering

is better than an act that merely leads to the relief of suffering, and an act that

is motivated by a disvaluing of suffering is better still Similarly, a belief thataims at the truth is better than one that merely leads to the truth, and onethat is motivated by a valuing of truth is better still Arguably, a belief that is

⁸ See DePaul (1993: ch 2) and (2001), and Zagzebski (1999a), expanded and reprinted in Axtell

(2000).

motivated by a valuing of truth or a disvaluing of falsehood has the value thatmakes knowing better than mere true believing Pursuing the belief/act analogy, Zagzebski concludes that true believing is not an end state analogous

to the relief of suffering Rather, true believing is an intellectual act, or at least,

it is strongly analogous to an act

The issue of how the reliabilist can handle the value problem is the topic of

Ernest Sosa’s essay, ‘The Place of Truth in Epistemology’ Suppose we think that

knowledge is belief that is both true and derives from intellectual virtue,where what makes a psychological feature an intellectual virtue is the reliabletendency of that feature to give rise to true beliefs If we also assume thatknowledge is more valuable than mere true belief, where does the value ofknowledge in addition to truth come from? Sosa offers an answer to this question that retains the idea that truth is the only fundamental epistemicvalue (with some qualifications for values such as understanding that are notdirectly connected to knowledge) Sosa proposes that we prefer our own successes, epistemic and otherwise, to be attributable to our own doing, andthis value can be intrinsic as well as instrumental Furthermore, there is alsowhat he calls ‘performance value’, the value of a belief performance thatwould normally produce true belief when operating in a suitable environment

A performance can have this value even when the ensuing belief is false Thechief intellectual goods involve hitting the mark of truth through the quality

of one’s performance

In ‘How to be a Virtue Epistemologist’ Christopher Hookway begins his reflections

with the schematic characterization of virtue epistemology as ‘approaches tothe most central problems of epistemology which give to states called “intel-lectual” or “epistemic” virtues a central or “primary” explanatory role’ Thischaracterization contains three elements that require comment: the centralproblems of epistemology, the nature of epistemic virtue, and the explanatoryprimacy of virtue Hookway addresses each of these elements, but what areperhaps his most interesting reflections concern the central problems of epistemology He points out that standard versions of virtue epistemologyaccept the typical contemporary view of the central problems, i.e that theyare to analyse the concepts of knowledge and justification and address scepticalchallenges by showing that it is possible for us to know, or at least have justifiedbelief Given this agreement with the rest of contemporary epistemologyregarding the central problems, virtue epistemology is distinguished fromother epistemologies only by the claim that the concepts of knowledge andjustification must be analysed in terms of virtues The acquiescence of most

virtue epistemologists in the consensus view of the central problems stands incontrast to the position that has driven many contemporary advocates ofvirtue ethics Virtue ethicists have tended to reject the contemporary consensusthat the central problems of ethics concern the moral ‘ought’, arguing that weshould instead concentrate on what is required to live well Is there space forvirtue epistemologists to mount a similar challenge? Hookway aims to showthat there is, arguing that instead of focusing on static states such as belief andthe evaluation of these as justified or knowledge, we might instead focus onevaluating and regulating the activities of inquiry and deliberation and therole of virtues in such evaluation and regulation.

Wayne Riggs proposes an alternative to standard truth-directed,

success-oriented epistemological theories in ‘Understanding “Virtue” and the Virtue

of Understanding’, arguing that the highest epistemic good is a state thatincludes much more than the achievement of true beliefs and the avoidance

of false beliefs In fact, it includes much more than knowledge: it requiresunderstanding of important truths So one way in which contemporary epistemology has been too limited is that it has focused on a less worthy goalthan the highest epistemic good Some of the intellectual virtues are bestunderstood as directed at understanding rather than at truth or knowledge.Intellectual virtues are also usually construed as traits that require reliablesuccess in reaching their goal, but Riggs argues that whether the goal is truth

or understanding, reliable success cannot be necessary for intellectual virtuesince some of the most intellectually virtuous persons, intellectual giants such

as Aristotle, Newton, and Galileo, are not noted for their success The lectual virtues should therefore be understood in terms of the values at whichthey aim, not at the values they reliably bring about When we give up truth-directed, success-orientated approaches in epistemology, the importance ofintellectual virtue becomes much clearer

intel-Christine McKinnon argues in ‘Knowing Cognitive Selves’ that the standard

epistemological requirements of impartiality on the part of the knower andpassivity on the part of the thing under investigation exclude from thepurview of epistemology a very important kind of knowledge: knowledge ofpersons Feminist philosophers have focused on problems in explaining

knowledge of other persons, but McKinnon suggests that the same

considera-tions require a reorientation in the way we think of knowledge of ourselves Inthis case the subjectivity of the knower is necessarily implicated, and thereflexive nature of the investigation means that what is known is unlikely toremain unaffected by the inquiry Justifying the knowledge each of us has of

our own selves poses enormous challenges to epistemology These challengescan be met if we see methods of acquiring knowledge and justifying claims toknow ourselves as continuous with the methods of acquiring and justifyingour knowledge of other persons Both are imbedded in social practices andboth involve mastery of a theory and responsible exercise of certain cognitivecapacities There are asymmetries between first-person and third-personknowledge, but these asymmetries neither rest on traditional claims of first-person privileged access nor do they undermine the possibility of knowingothers The project of coming to know persons is a project of coming to knowtheir moral and cognitive characters The case of self-knowledge highlightssome interesting points of intersection between virtue ethics and virtue epistemology and may illuminate some methodological issues in contemporaryepistemology.

The most interesting parts of works from the virtue ethics tradition areoften the detailed, perceptive treatments of specific virtues and vices Ourhope is that contemporary virtue epistemology will eventually produce similarly rich discussions of intellectual virtues and vices In ‘Humility and

Epistemic Goods’, Robert Roberts and Jay Wood provide a model for the kind of

discussions we hope to see They begin their treatment of intellectual humility

by examining the broader, moral conception of humility Their strategy is tosituate humility in relation to its various opposing vices, which include arrogance, vanity, conceit, egotism, grandiosity, pretentiousness, snobbishness,impertinence, haughtiness, self-righteousness, domination, selfish ambition,and self-complacency Roberts and Wood focus on vanity and arrogance inparticular They characterize vanity as an excessive concern with how one isregarded by other people and arrogance as a tendency to infer illicit entitle-ments from one’s supposed superiorities Humble as opposed to vain peopleare unconcerned with and inattentive to how they appear to others This does not mean that humble people are ignorant of their good qualities, justthat they are not particularly interested to be recognized for having thesequalities The reason for this is that their attention is focused on other, moreimportant things In the case of intellectual humility, one such thing wouldtypically be the truth Thus, for example, while vain persons might seek tohide their errors for fear of what others might think of them, the humble will

be more concerned that any mistakes be brought to light so that they can correct their errors and get their inquiries back on track Humble personsare not distinguished from arrogant persons by being unaware of or evenunconcerned with entitlements The distinction turns on what motivates the

awareness or concern Paradigmatic cases of arrogance involve an excessiveinterest in entitlements motivated by what Roberts and Wood call their ego-exalting potency In contrast, when humble people do have an interest insome entitlement, the interest is pure, in the sense that they are concernedwith the entitlement because it serves some valuable purpose or project.Roberts and Wood close their essay by considering a wide variety of ways in whichintellectual humility promotes the acquisition of epistemic goods.

Over three decades ago Roderick Chisholm observed that ‘many of the characteristics which philosophers and others have thought peculiar to ethicalstatements also hold of epistemic statements.’⁹ These days we may be lessinclined to focus on the linguistic form in which ideas are expressed than onthe ideas themselves, but Chisholm’s point still holds Much of what moralphilosophers talk about applies to epistemology, although epistemologists andethicists usually formulate the problems of their respective fields differently.The problems of epistemology have evolved over the last few decades and thedispute between foundationalism and coherentism no longer dominates the field Sosa’s suggestion that the idea of an intellectual virtue can illuminatethat dispute is no longer the main attraction to virtue in epistemology Theintroduction of the idea of virtue into epistemological discourse has led to

a new set of problems and issues for discussion in epistemology that overlapwith value theory A number of new directions for research are suggested bythe chapters in this volume and we hope that this book will encourage furthercollaboration between virtue ethicists and virtue epistemologists

⁹ Chisholm (1969: 4).

Part I

Classical Virtue Ethics and Virtue

Epistemology

Virtues of the Mind,¹ which has opened up many exciting paths of researchexploring the links between virtue ethics and epistemology This is a grippingand seminal book, which will surely change the contours of its field, and bringtogether two areas which have functioned in mutual isolation and can onlygain from the discovery of their links We all owe Zagzebski thanks for her pioneering work and its effects My own contribution comes from the direction

of virtue ethics, and I shall be exploring two aspects of the structure of virtue,

as that has developed in the ancient virtue ethics tradition, which have ations for the relevance of virtue to epistemology I shall have less to say about the details of the application, since epistemology, at least modern epistemology, is not my area of specialization; but I am fairly confident thatthey are central to the project of using a rich notion of virtue to illuminateepistemological issues

implic-¹ Zagzebski (1996) I shall also refer to Zagzebski (1999b) and to the exchange between her and various commentators in the Book Symposium on Virtues of the Mind in Philosophy and Phenomenological

Research: Zagzebski (2000a) and (2000b), Greco (2000a), Alston (2000), Kvanvig (2000), Kornblith

(2000), and Rorty (2000) Zagzebski usefully distinguishes her work from previous work in logy which featured terms such as ‘intellectual virtue’ or ‘virtue epistemology’ but which made

epistemo-no appeal to the epistemo-notion of virtue as that has been developed in the virtue ethics tradition.

The issues I shall focus on are those of virtue and skill, and virtue and success.

In both cases we get a clearer picture if we look at the whole ancient virtuetradition, rather than emphasizing Aristotle For contingent historical reasonsAristotle’s has been the theory on which most philosophers focus when theyturn to virtue.² But treating Aristotle as authoritative for virtue ethics fails to

do justice even to the ancient tradition For hundreds of years different theories were proposed within the framework of happiness and virtue, andthere was extensive inter-theory debate As a result, we can separate theframework and main assumptions of virtue ethics from the specificities ofAristotle’s own theory Sometimes this can turn out to make a large difference

as to what is implied by the use of a ‘virtue ethics’ approach, and I shall bearguing that for these two issues it does In both cases, if we look at the wholevirtue tradition, we find important implications for the relation of the moral tothe intellectual virtues, and, hence, for the relation of ethics to epistemology

II

Aristotle rejects the idea that virtue is a skill (Virtue here is moral virtue, asindeed is standardly assumed in ancient ethical discussion;³ we shall get tointellectual virtue shortly.) This may strike us as unsurprising, indeed merecommon sense But it is significant that Aristotle is a lone voice here Theancient virtue ethics tradition followed Plato and the Stoics in holding thatvirtue is a skill That is, it is a kind of skill, there being other kinds as well;virtue is, as the Stoics put it, the skill of living The claim that we should follow the ancient tradition rather than Aristotle may at first sound rather

² For one thing, Aristotle’s lecture notes on ethics have come down to us in a more complete form than have those of other ancient schools like the Stoics and Epicureans (This has not been

an unmixed blessing, however, since the Nicomachean Ethics (hereafter NE)—though not, ingly, the Eudemian Ethics—has been treated as though it were a continuous production, like a

interest-modern book, rather than a collection of notes, sometimes with differing treatments of the same issue.)

³ Greco (2000a: 180–1) argues that virtue is a wider notion than that of moral virtue, covering the idea of excellence in general In her reply, Zagzebski (2000b: 207–8), is inclined to think that

the issue is merely verbal, and hence also the issue of whether reliabilist theories can be regarded

as a type of virtue epistemology However, within the ancient tradition of discussing virtue it was assumed that, while there was a broader use of ‘virtue’ to mean any excellence, the more proper use of the word was to apply it to moral virtue For reference to passages making this point see Annas (1993: 129–31).

academic, but this issue of whether virtue is or is not a skill is not merely ofhistorical interest: it raises philosophically crucial issues about the intellectualstructure of virtue.

Aristotle says that there are many points of difference between a virtue and

a skill This is obvious enough, and modern writers have developed and modified the list of differences which he sets out.⁴ For example, skill involves

a mere capacity, can be forgotten, and is less precise than virtue is.⁵ However,

we can reasonably ask how much such differences matter: the thesis thatvirtue is a skill is a claim that virtue is one kind of skill, and thus that the idea

of skill is central in helping us to understand what virtue is Against this claim,pointing out obvious differences between virtues and skills is ineffective.⁶How can virtue be a kind of skill? It has the intellectual structure of a practicalskill Even Aristotle recognizes this; it enters right at the start of his account ofvirtue, where he compares learning to be just with learning to be a builder.You have a role-model, and first you copy what he or she does, then come tounderstand for yourself what the point is of doing what that person does.Increase of understanding goes with increased autonomy of reflection andaction This is why, for both skill and virtue, you need a teacher to begin with,but then become able to act on your own independently

Underlying this simple fact is a connected set of epistemologically interestingpoints about skill First, a skill or expertise is teachable There is some intellectualcontent to be conveyed, not just picked up by external mimicking Wherethere are teachers and learners, we have something which at least in principle

is an expertise, not just a matter of an empirical ‘knack’ to be picked up.Second, the expert is someone who has an understanding of her subject mat-ter as a whole This is a demanding condition Someone learning, say,

a language, will pick up bits of the subject here and there—the future tense,vocabulary, and so on The expert in the language will have mastery of all that

⁴ See e.g Wallace (1978, 1988) Zagzebski follows Wallace fairly closely in (1996: 106–16).

⁵ Skill involves a mere capacity, for knowledge can be used in ways opposed to the right ones,

whereas a virtue cannot be used equally well for opposed ends (NE 1129a11–16); virtue cannot be

forgotten, whereas skills, being mere intellectual states, can (NE 1140b 28–30); with a skill the son who deliberately makes mistakes is preferable to the person who makes them without intend-

per-ing to, whereas the reverse is true with virtue (NE 1140b 22–8); virtue is more accurate than skill is (1106b14–17).

⁶ For a fuller discussion see Annas (1993: ch 2, sect 3, esp 67–84; also Annas (1995) Aristotle

makes the claim (NE 1105a26–b5, 1140a2–6, 16–17, b1–4, 6–7) that with skills all that matters is that the product be good, whereas with virtue actions cannot be appropriately judged without bring- ing in the agent’s intentions This is striking, but, as we shall see, not decisive.

is needed to understand the language, and, moreover, will see how it is all unified Similarly, someone learning a practical skill like building will pick upbits of know-how and technique here and there; the expert, however, willhave mastery of everything relevant to that kind of building, and will haveunified that mastery so as to be able to understand his own and others’ successes and mistakes, and to be able to apply his skill in new situations with-out further learning being required And thirdly, an expert is able to articulate

her understanding of her subject, able to ‘give an account’ of it, logon didonai, in

the ancient way of looking at it She is able not only to unify the various ments she makes within her field, and the actions she does, but to explainthem and, if necessary, justify them, in terms of whatever general principlesare needed to express understanding of the subject

judge-These conditions are not independent of one another, since a teacher canscarcely teach if she is completely inarticulate about her subject, and what istaught must be a unified body of practical knowledge rather than a bunch ofunconnected practical tips, if it is an expertise that we have.⁷

How is virtue a skill? It shares the intellectual structure of a skill in thesethree ways This is visible in the account Aristotle gives of the acquisition ofvirtue,⁸ but it can be seen most lucidly in the Socratic dialogues of Plato.There, Socrates challenges people who appear to be experts about somevirtue, such as courage, but fail to unite their isolated beliefs⁹ and to offer anyarticulate unified understanding of the matter; this shows that they lackunderstanding of the virtue in question Laches, in the dialogue named afterhim, can give examples of courage by pointing to men fighting in battle But

at first he can provide no account at all of what this kind of action might have

in common with other kinds of brave behaviour—coping bravely with illness

or poverty, for example—and when he does, his suggestion is obviously less at explaining how and why all the very diverse kinds of brave action arebrave Laches, who is supposed to be an expert in bravery, has failed to convey

hope-⁷ I have found that it is common at this point for some people to object that we ically recognize kinds of expertise where these conditions are not met I cannot go fully into the matter here, but I believe that, while these conditions are demanding, they are not alien to our intuitions, and that on reflection we do in fact deny that someone is an expert if she is inarticu- late about her subject, unable to teach it, or unable to express more than isolated tips about its practice ⁸ See Annas (1993: 67–69).

commonsens-⁹ Which may be correct; Socrates is not implying that Laches, e.g., in the Laches, is wrong about

the kind of act which is brave Where he fails is in having any unified understanding of courage;

he fails the conditions for being an expert.

any articulate understanding of bravery; clearly he is, despite appearances, noexpert in it The same undermining of claims to expertise in moral mattersoccurs in other dialogues, and Plato scholarship has for some time now recognized that the kind of knowledge or understanding which is requiredand found lacking is the kind of knowledge that an expert possesses.¹⁰ Indeed,Socrates is always appealing to practical skills such as those of the navigator,doctor, or farmer to illustrate the kind of practical understanding that heseeks in moral matters Nor is this concern peculiar to Plato The Stoics takeover the thesis that virtue is a skill and develop it explicitly and at length, forthe same reasons which appear in the Socratic dialogues.¹¹ It became so standard in ancient ethical theory that it could be taken for granted in anyserious debate between ethical theories.

There are common objections at this point from modern philosophers.Kinds of practical expertise have ends which are fixed; it is clear and uncon-troversial what counts as success in a skill like navigation, or car repair Andthese ends are conditional in their hold on our motivation; our interest inexercising skills depends on our concern to obtain their ends Can it then bereasonable to think of virtue as having the structure of a practical expertise,given that the end that virtue aims to achieve—a well-lived life—is one which

is neither clear nor uncontroversial, and also one from which we cannot similarly become motivationally detached?

The answer to this in the ancient virtue ethics tradition is clear, and best put

by the Stoics (though it can also be found in Plato) Virtue is ‘the skill in living’, and living your life is an end which everybody has, and which, short ofsuicide, is non-detachable By the time you start to reflect about your life andthe best way to live it, you already, as we put it, have a life You already have afamily context, for example, and a socio-economic context, with some kind ofemployment and income You already are the product of some kind of education,including moral education, and have certain values and priorities For theancient virtue tradition, all this is your raw materials, on which you get towork as you develop virtue, aiming to make your life a product of under-standing rather than conformity, something unified around pursuit of goodvalues rather than driven by isolated desires or run by the values of others.Virtue, then, is a global expertise in your life, and will always differ from localkinds of expertise in just these two ways, namely that the end we seek in

¹⁰ See Woodruff (1990) for good discussion of the issue.

¹¹ For the Stoics see Annas (1993: 69–70).

becoming virtuous is not antecedently fixed in the way that the end of carrepair is, and also that living our life is not an end that we can cease to careabout, as we can cease to care about having the car fixed We can choose, ofcourse, to live our lives in a thoughtless and random way rather than to livethem in a way which tries to improve them in the light of unifying under-standing; but this does not make the end of virtue detachable in the way theends of local skills are.¹²

Moral virtue, then, is a skill in the ancient virtue tradition; it is an expertise,

a kind of practical knowledge Local, mundane skills serve as examples of thekind of unified practical understanding which, if we become virtuous, willorder our lives in a unified way based on understanding

How, then, does moral virtue relate to intellectual virtue? The right answerhas been elegantly stated recently by Paul Bloomfield.¹³ Moral virtue is onekind of skill, intellectual virtue is another Moral virtue, as explained, is a kind

of practical knowledge which is illuminated by practical kinds of expertise.Intellectual virtue is another kind of skill Neither should be seen as a sub-kind

of the other—although of course any realistic account of the moral life willfind many complex connections between them.¹⁴

It is only to be expected that intellectual virtues should have a strong lectual structure unified by understanding Must an intellectual virtue, however,have the same intellectual structure as the moral virtues that we have seensharing the structure of practical skills? If expertise is marked by the three conditions discussed above, of teachability, unified understanding of the field

intel-as a whole, and articulate ability to give an account of what is understood,then intellectual and moral virtues will share this structure Intellectualvirtues, however, appear to be more various in their structure than moralvirtues are, in a way that doubtless owes something to the fact that theoreticalskills are more various in their structure than practical skills are Aristotle’s

intellectual virtues in Nicomachean Ethics VI are highly diverse The virtue aiming

¹² Zagzebski (1996) is rather quick to dismiss what she calls a ‘happiness-based’ version of virtue ethics on the ground that the notion of happiness in question would have to be fixed and based

on now unacceptable teleology (e.g pp 201–2) The constraints on eudaimonist theories are formal ones, and they allow for considerable rethinking of the aim of happiness in the light of the demands of the theory of virtue This issue is discussed, with the ancient evidence, in Annas (1993:

pts 1, 4), and also in Annas (1998a). ¹³ In Bloomfield (2000).

¹⁴ Cf Zagzebski (1996: 158–65) Zagzebski is also right that Aristotle’s own discussion in NE VI

does not give an adequate line of distinction between intellectual and moral virtues—although he does at points indicate where there might be conflicts.

for demonstrative knowledge, for example, is different in structure from thevirtue aiming for non-demonstrative knowledge The same goes for the Stoicsubdivisions of wisdom, and for Plato’s collection of intellectual virtues.¹⁵ Thestructure of an intellectual virtue will naturally depend on the scope and type

of the relevant intellectual skill; it would seem that we might have several differently structured intellectual virtues which all met the conditions forexpertise

It could be argued that the moral virtues essentially involve emotions andfeelings in a way not true of the intellectual virtues Indeed, moral virtue crucially involves in its development the progressive control and finally trans-formation of the person’s emotive side.¹⁶ But it would be a mistake to holdthat development of an intellectual virtue like perseverance or intellectualhonesty never involves such control and transformation of recalcitrant, notpurely intellectual, elements of the person Moreover, development of theintellectual virtues may straightforwardly require such transformation of theemotions and feelings by way of the development of a moral virtue Honesty

in some research, for example, requires that the person not be under the ence of greed for money; indeed, honesty seems to be the same moral virtuewhether applied in handling money matters or in conducting research.The real distinction emerges when we consider that moral virtue is essentiallypractical; it is the skill of living, where living, in the virtue tradition, is seen asessentially active, shaping your life so that it is ordered from within The wayyou live is seen as actively reflecting and expressing your character and henceyour choices Intellectual virtue, on the other hand, is not essentially practical;

influ-it is theoretical in that influ-it is directed at achieving aims other than good action.Particularly if we think of intellectual virtue as aimed at achieving truth, wecan see that its aim is going to be distinct from that of moral virtue

Of course, there might still be a close connection between the two kinds ofvirtue, and most virtue theories have thought that there is One view frequently found attractive is that the intellectual virtues, whose aim is truth,

¹⁵ At Republic 487a (and cf 490c) we find that the ideal person to achieve knowledge, in ideal

conditions, must, as well as having a good memory, be good at learning, large-minded, and

‘eleg-ant’ (eucharis, probably meaning that he presents himself and his work in an attractive rather than

harsh or gauche fashion) He or she must also be attracted to and ‘akin to’ truth, as well as ing the moral virtues of justice, courage, and temperance This is a collection of very different intellectual virtues There would not seem to be any a priori reason why being large-minded, with

hav-a brohav-ad vision, should be structurhav-ally like hhav-aving hav-a drive to discover truth.

¹⁶ See Annas (1993: ch 2, sect 2).

deepen the understanding which is the basis of the moral virtues After all, the

moral virtues are aimed at doing the right thing, and this can scarcely allow

indifference to the truth of your beliefs about the matter Even if the intellectualvirtues enable us to discover truths about matters which are recondite andabstract, still our increased grasp of truth will serve to broaden and deepen theunderstanding at the basis of the kind of practical knowledge which is moralvirtue As Zagzebski puts it, ‘[I]f it turns out that the ultimate end of truth and

the ultimate ends of the moral virtues are all components of a life of eudaimonia, then the moral and intellectual virtues do not even differ in their ultimate

ultimate ends.’¹⁷

What, though, if this turns out not to be the case? In the virtue traditionthere are two conflicting lines of thought on this, both of which are foundappealing by both Plato and Aristotle While they think most of the time thatseeking truth will form part of a life well-ordered by moral virtue, they both

at some points express a contrasting thought: seeking truth can become anend indifferent to or even conflicting with the end of living according tomoral virtue.¹⁸ The attractiveness of the intellectual search for truth, and theintrinsic appeal of its objects, can lead humans away from the aim of living

a morally ordered life It can lead them to aim to devote their energies entirely

to the search for truth, to the point of wishing to transcend the boundaries ofhuman life altogether and to try to ‘become immortal’, as Aristotle famouslysays in this connection In this case the pursuit of happiness in a morally unified life will have been disrupted Someone who seeks truth in a way which

is indifferent to or conflicts with living a morally virtuous life is still, however,exercising the intellectual virtues It is unconvincing to claim that someonewhose intellectual pursuit of the truth conflicts with leading a moral life mustreally be lacking in intellectual virtue (Indeed, it is likely to be the intellectuallyvirtuous achievers, rather than the intellectually faulty, who have this problem.)The intellectual virtues can, though they need not, have a differing aim fromthe moral virtues, since the theoretical aim of truth can come into conflictwith the aim of moral virtue, which is a practical type of knowledge

The relation between intellectual and moral virtue that emerges from thevirtue ethics tradition, at least in its developed ancient form, is that both are

¹⁷ Zagzebski (2000a: 173).

¹⁸ Plato expresses this memorably: see the picture of the philosopher’s indifference to ordinary life

and its virtues and vices in the ‘digression’ in the Theaetetus, and the way the ‘Guardians’ are forced to rule in the central books of the Republic In Aristotle there is the well-known conflict between the body of the NE and the passage that has come down to us as the second part of ‘Book 10’.

kinds of skill or expertise, whose aims can but need not converge Takingvirtue seriously in the epistemological framework of the intellectual virtues,then, does not give support to thinking of intellectual virtues as a subset ofmoral virtues, nor to taking epistemology to be properly subsumed underethics.¹⁹ Taking both kinds of virtue seriously, however, may be fruitful inother ways For one thing, taking moral virtue seriously reveals how intellec-tual a structure it has, and this suggests that virtue ethics might get aid fromepistemology, as well as epistemology benefiting from virtue ethics Ethics andepistemology can produce mutual benefit from mutual study.

III

The second issue is that of virtue and success Zagzebski frequently insists that

‘virtue is a success notion’,²⁰ and in this she and others are following not only

Aristotle, who insists that the virtuous person is successful (katorthotikos), but

the Stoics, who call a virtuous action as performed by a virtuous person

a success (katorthoma).²¹

The success element in virtue is important for anyone wishing to develop

an epistemology in which virtue plays a basic or foundational role For ledge is a success term if any is Knowledge is not the state you achieve bydoing your best though you fail, but the state in which you actually succeed

know-in gettknow-ing your claim right, and succeed know-in meetknow-ing the required conditions,

¹⁹ As Zagzebski (1996: ch 7) claims Her arguments, however, do support many weaker claims, for example that epistemology is more closely connected to ethics than many recent epistemo- logists and ethicists have thought Epistemologists freely use ethical notions in developing their theories, and theories in epistemology frequently mirror ethical theories in their structure; Zagzebski is surely right that this should be done in a self-conscious and careful manner I would add that writers in ethics have frequently had to develop a moral epistemology in isolation from modern developments in epistemology, which have focused on morality and moral epistemology only from their own perspective.

²⁰ Zagzebski (1996: 136–7, 176–84, 1999b: 107, 2000a: 174–5) At (2000b: 211) she responds to the

objection of Alston and others that virtue may not be so closely connected with success with the irenic suggestion that reasonable people may differ as to the importance of the success element in moral assessment.

²¹ Aristotle, NE 1104b 34, where the good person is said to be successful about the fine, the advantageous, and the pleasant, the three sources of motivation for choices and avoidances.

Katorthoma in Stoic texts is usually translated by a different term from ‘success’ in English, to avoid

confusion; thus Inwood and Gerson (1997) use ‘(morally) perfect action’ Long and Sedley (1987) translate it, rather weakly, as ‘right action’.

whatever these may be Zagzebski defines knowledge as a state of cognitivecontact with reality arising out of acts of intellectual virtue,²² and obviously

‘act of virtue’ must be a success term here, or it would not be knowledge that

we were defining Virtue, then, must be a success term in virtue epistemology.When we look at the virtue ethics tradition, however, we find that the relation of virtue and success contains complexities, and that when these areexamined we find that we must also introduce complexities into any attempt

to make use of virtue in reaching a definition of knowledge

The virtuous person must have the right motivation, and must also reliablysucceed in what she does An act which fails to achieve its aim can be said to

‘lack[s] something morally desirable’.²³ But what is the virtuous person’s aim

in acting? She has two One is her telos or overall aim, of living virtuously and

acting from motives of virtue.²⁴ Virtue, after all, is a settled state of the person,with the overall aim of making the person’s life as a whole be one way ratherthan another, virtuous rather than evil or complacent (Living virtuously, further, either constitutes, or contributes to, happiness; but that is a distinct

issue.) The virtuous person’s other aim is what the Stoics call her skopos or

immediate target, which is what is aimed at in any particular case of acting virtuously The target of a just distribution will be everyone’s getting whatthey are entitled to, that of a brave rescue will be the safe conveyance of peopleout of the burning building, and so on

Plainly, someone can succeed in achieving the immediate target of anaction on a particular occasion without achieving the overall aim of living virtuously This will be the case if the person is not virtuous, and so does anaction which is the kind of action which a virtuous person would characterist-ically do, but does not do it as a result of the virtuous person’s motivation.²⁵Equally plainly, a virtuous person can succeed in achieving the overall aim of

²² Zagzebski (1996: 270–1, 1999b: 108–9, 2000a: 174–6) In all cases Zagzebski gives an alternative

definition in terms of belief, which is narrower and links more directly to traditional definitions

in terms of propositional belief. ²³ Zagzebski (1999b: 107).

²⁴ We should note that it is a mistake (often made) to think of this as egoistic Her aim is to be

an honest person, that is, to give others their due, to think of herself precisely as standing in moral

relations to others This is not egoistic, and has nothing to do with the condition of thinking that

it is the state of your own character which matters, rather than other people and what you owe

to them It needs to be emphasized that ancient theories of virtue are not focused on the self rather than on making the world a better place.

²⁵ In such a case the Stoics say that the action is an appropriate action or kathekon, which is

defined as an action such that you can offer a reasonable defence of having done it Only an action

done by a virtuous person from the right motivation is a katorthoma.

living virtuously by performing a virtuous act, even if, through no fault of herown, she fails to achieve the immediate target If the brave rescuer does every-thing he can, takes the appropriate precautions, and so on, but the victims dieanyway because they are shot on the way out by a deranged gunman whohappens to be there, then the brave action has failed to achieve its immediatetarget, but not in a way which implies that the brave person has failed toachieve his overall aim of living virtuously and so acting, in this, case, bravely.

It is crucial, therefore, in examining a virtuous act, to ask what kind of success

is in question—success in achieving the overall goal or success in achievingthe immediate target For achieving the overall goal is a matter of having theright motivation (something, of course, which in a virtue ethics is the result

of a lengthy and demanding process), and this is up to the agent, since it is shewho makes her life be one kind of life rather than another But success inachieving the immediate target may not be in this way up to the agent, andmay depend on various kinds of moral luck

From a virtue ethics point of view, which is the success that matters? Virtueethics is concerned with the person’s life as a whole, with character and thekind of person you are The right perspective on an action, therefore, will forvirtue ethics be the one which asks about success in achieving the overall goal,rather than success in achieving the immediate target What matters is whatthe person’s motivation was, and how this relates to her developed characterand life as a whole; for this is her achievement, what she has made of her life

To the extent that success in achieving the immediate target depends on tors over which the person has no control—moral luck of various kinds—itwill be of less interest to virtue ethics Success or failure in achieving theimmediate target will affect various judgements we make about the action,but if, like the Stoics, we distinguish clearly between the immediate target andthe overall aim, it is achieving the latter, not the former, which will make the

fac-action a success, a katorthoma Here virtue ethics parts company with theories

like (most forms of ) consequentialism, for which it is the actual results thatmatter for our evaluation of the agent,²⁶ and stands with Kantianism, forwhich what matters is the agent’s motivation.²⁷

²⁶ Modern versions of virtue ethics which align it with forms of consequentialism are thus abandoning the ancient virtue ethics tradition This is perfectly reasonable, though it would avoid confusion if such theories made it clear that they are talking about a different sense of virtue from that found in the virtue ethics tradition.

²⁷ Though virtue ethics, because of its focus on the agent’s character and life as a whole, has a richer conception of the agent’s motivation than forms of Kantianism which focus on motivation

at the time rather than on more established states like the virtues.

Virtue is a skill, in the virtue tradition, the skill of living your life in a waywhich turns your raw materials into a life lived with and from understanding.

It is a global skill, as we have seen, and this explains why it is compatible withfailure to foresee some particular circumstances Our ordinary notion of skillgoes some way towards making this point; we sometimes judge an expert bythe exercise of her skill rather than by the product, as when an expert musicalperformer produces a better performance on a bad instrument than a lesserexpert can on a good instrument, even if the latter can be said to sound better.Still, the Stoics insist that virtue is not to be judged by its actual results, like

a skill which is judged by its products, but is more like performance skills, such

as acting or dancing, where the excellence that is judged is the excellence inthe activity and not in some separable result For with virtue it is not theresults which define success: ‘actions initiated by virtue are to be judged rightbeginning from their first inception and not in their completion.’²⁸

This is an issue where privileging Aristotle can lead to confusion, since onthis point he is confused On the one hand, he insists on the praiseworthiness

of virtue, and the importance of choice as opposed to action in distinguishingcharacters.²⁹ But on the other hand he also stresses success in achieving thetarget in the practice of various virtues, sometimes in cases where this isexplicitly not up to the agent, as with the ‘virtue’ of magnificence (the ‘virtue’

of spending money on civic projects), which only a rich person can exercise.³⁰There is an unresolved internal tension in his theory as a result, mirroring hisuncertainty as to the role of external goods in the virtuous life generally.³¹ Wecan see the problem if we ask about the role of a ‘virtue’ like magnificence Toexercise it the person needs not only external goods, in this case money, whosepossession is a matter of moral luck; he needs to be actually successful in hisexercise of tasteful spending, producing what succeeds in impressing the audi-ence without overwhelming them, and so on Aristotle, however, also believes

in the mutual reciprocity of the virtues; to have one you have to have themall But he clearly does not believe that if you are fully brave, you have all theother virtues, and therefore have magnificence, and therefore have magicallyacquired lots of money and taste This example shows that, since the virtues

²⁸ The spokesperson for Stoicism in Cicero, On Moral Ends (De Finibus) III 32 See Annas (1993:

403–5) for Stoic theses about virtue, skill, and success. ²⁹ NE 1111b

are mutually reciprocal, none of them can depend for their exercise on moralluck; hence magnificence, which does so depend, is not a real virtue Aristotlefails to draw this conclusion because he is too respectful of conventional viewswhich think of the activities of rich people, like magnificence, as virtues.

On this issue the Stoic view is much clearer and more defensible thanAristotle’s Of course it is often not up to me whether my action achieves the immediate target; but is it up to me whether I succeed or fail in acting virtuously—that is, with the right motives, from a developed disposition andwith the right reasoning? If it is not, then it is not up to me whether or not

I can become a moral person; and the Stoics are not alone in finding this anunacceptable position.³²

It is sometimes urged that we feel more admiration for the act which, aswell as being virtuous and thus succeeding in the overall aim, also actuallydoes get its target; that this is the sense in which the act that fails here ismorally lacking But this seems not to be true Take Socrates’ defence speech,

the Apology,³³ in which he uncompromisingly defends the values that he has

lived by, and refuses to pander to the jury’s values even at the risk of being executed for not doing so Do we admire Socrates less because in fact he failed

to swing the crucial thirty votes?³⁴ Do we think of him as a pathetic loserbecause he failed to express the degree of deference to the jury that wouldhave secured his acquittal? Surely, rather than finding his action morally lacking,

we admire him all the more for refusing on this occasion to compromise hisvalues—if anything, his knowing refusal to do what was required to secure his immediate target makes us more convinced that he succeeded in achievinghis overall aim of living a virtuous life

It can be suggested³⁵ that we continue to praise the agent, but give the act lesspraise This distinction can do work in some kinds of ethical theory, but in avirtue theory is problematic For the suggestion here would be that we praise

Socrates for being virtuous, living the life he does and having the character he

does, but on this occasion we fault his action Why, however, do we fault it?For being so uncompromising as to lead to failure in worldly terms But this is

³² There is, of course, a large debate here over the relative merits of the Aristotelian position versus the Platonic–Stoic one as wholes, and the entire issue of moral luck and its roles, if any, in ethics What is relevant here is the point that the Stoics have much the stronger and more defens- ible position on virtue, skill, and success.

³³ I am here talking about Plato’s version; Xenophon’s Apology raises quite different issues.

³⁴ The jury consisted of 501 citizens; Socrates was condemned by a majority of 60.

³⁵ See Zagzebski (1996: 136–7, 1999b: 107, 2000a: 174).

to say that we fault it for being just the kind of act which this kind of personwould do! To fault what Socrates did for its lack of success in achieving the target precisely is to fault Socrates for being the person he is, and for actingaccordingly Of course we can deplore the actual results of the action on thisoccasion Virtue ethics can account as well as other theories for the fact thatoften we wish that the world had gone well and been improved in a way thatdid not happen.³⁶ We, as well as Socrates, can regret that the thirty votes wentthe wrong way; there is no reason to think that virtue ethics is more indiffer-ent to the results of actions than other theories are But the relevant pointcannot be put, within a virtue ethics, by separating assessment of the agentfrom assessment of the act In so far as it was a virtuous act, done by a virtuousperson for reasons of virtue, it cannot be faulted from the virtue point of view.

It is the jury we wish had been different, not Socrates’ action

It is also doubtless true that virtue is generally reliable in producing success

in getting the immediate target As Terence Irwin puts it:

It is easy to see why, in favorable external conditions, virtuous people will have moreobjective success than other people will have For they will have done all that can reasonably be expected of them; and if they do that, they will have tried to find all therelevant information that they could reasonably be expected to find, taken propercare, and so on It is not surprising that action on these principles will often result inobjective success.37

A virtue ethics approach can take all this into account; but when the virtuousperson fails to get her target through no lack or fault of her own, a history ofusual success here is not to the point We have to choose which kind of successmatters, and any virtue ethics in which the issue is clearly faced comes down

on the side of success in achieving the overall aim, which is compatible withfailure to achieve the immediate target

How does this matter for the application of virtue theory to epistemology?

As I mentioned, knowledge is a success term, and so a theory which definesknowledge in terms that feature virtue must take virtue to be a success term.Knowledge will, of course, be defined in terms of intellectual, not moralvirtue, but for a theory which holds that these have the same structure or thatthe intellectual virtues are a subset of the moral ones this issue will be the

³⁶ Zagzebski (1996: 137) rightly stresses that ‘morality is also in part a project of making the world a certain kind of place—a better place, we might say, or the kind of place good people want

it to be.’ ³⁷ Irwin (1990: 71).

same So we must ask: what kind of success in intellectual virtue will berequired for a workable definition of knowledge?

Here it looks as though the answer we get from ethics about moral virtue

is the wrong one for epistemology We surely do not want to define knowledge

in terms of an overall disposition to succeed which is compatible with particularfailures due to ‘epistemic bad luck’ It would have the implication that I couldhave knowledge as a result of being intellectually virtuous, even though onthis occasion, through no fault on my part, I am in fact wrong A viable definition of knowledge in virtue terms must surely avoid this, and so must beaiming at success in achieving the immediate target I have knowledge if I amright, say the truth (and also meet some further conditions) But, as we havejust seen, this is not what a theory of ethical virtue demands; what is of mostinterest for it is success in achieving the overall goal It looks as though therequirement for virtuous success in ethics—achieving the overall goal—isprecisely what a virtue epistemology has to reject Similarly, the requirementfor virtuous success in epistemology—success in achieving the immediate target—is precisely what a virtue ethics has to reject

It is because of this point, I think, that problems arise for attempts, such asZagzebski’s, to define knowledge as a state arising from acts of intellectualvirtue What is an act of virtue, in this theory? It is defined as follows:

An act is an act of virtue A if and only if it arises from the motivational component of A,

is an act that persons with virtue A characteristically do in the circumstances, and issuccessful in bringing about the end of virtue A because of those features of the act.38

Zagzebski claims that ‘the concept of an act of virtue is something we wouldwant in an ethical theory anyway’,³⁹ apart from its application in epistemology.This may be true, but what I am now concerned with is the issue of how well

it fits a virtue-ethical theory in particular

First, you do not need to have the virtue in question to perform an act ofthat virtue in the technical sense.⁴⁰ In the virtue ethics tradition, this wouldmean that the act is not an act of virtue at all; it is just an ‘appropriate action’

³⁸ Zagzebski (1999b:108) The version at Zagzebski (1996: 248) adds a couple of complications which do not affect the present issue Zagzebski (2000a: 175) gives a definition which includes the

point, stated more generally, that ‘the end of virtue A includes the ultimate end of virtue A as well

as the proximate end.’ But, as we have seen, this point is problematic.

³⁹ Zagzebski (2000a: 176).

⁴⁰ Zagzebski (2000b: 209), in response to Greco, is unwilling to require that an agent

perform-ing an act of virtue have the virtue in question on the grounds that this would make the grounds for having a virtue too weak.

such as anyone can perform when trying to become virtuous An act

performed by a non-virtuous person could not be a ‘success’, a katorthoma, and

so could not be a starter as an act of virtue in any sense

The impact of this, however, is somewhat softened by Zagzebski’s requirementthat to perform an act of virtue you need to have the virtuous motivation inquestion, at least the motivational component of it, although you are not virtuous The idea is presumably that you have the virtuous motivation, but

do not (yet) have it in a sufficiently robust, reliable, and integrated way to bevirtuous Hence you can perform an act of virtue in the technical sense, withouthaving the virtue

Here the problem is that this is the wrong story about virtuous motivation.For all versions of the virtue ethics tradition, the motivation of the non-virtuous

person is different from that of the virtuous Aristotle says that the virtuous person is motivated by the ‘fine’ (kalon) while those who are not yet virtuous are

not—they are motivated by advantage or pleasure Becoming virtuous is not

a matter of already having a small amount of the right motivation which, so

to speak, spreads and grows bigger; it is a matter of learning to change Becoming virtuous is learning to acquire the right motivation; if you already

had it, even on a small scale, becoming virtuous would be easier than it is

A virtue, in the virtue ethics tradition, is a complex matter, and coming tojudge rightly and to be rightly motivated go in tandem, develop slowly andinvolve the person’s becoming responsive to considerations which precisely do

not resonate with the non-virtuous.⁴¹ Someone without the virtue in question,then, cannot possess its motivational component If an act can be performed

by a person without the virtue in question, then it can be performed withoutbeing motivated in the way appropriate to that virtue The technical notion of

an act of virtue tries to combine the point that it can be performed by one lacking the virtue in question with the point that it is neverthelessbrought about by means of that virtue, since it is to be brought about by thatvirtue’s motivational component But these two points cannot be reconciledwithin a virtue theory, at least one in the virtue ethics tradition

some-Second, an intellectually virtuous person, in this kind of virtue epistemology,may reason in a way that is virtuously motivated and follows the appropriateintellectually virtuous reasoning; but they may fail, because of epistemic badluck, to get the right result, and in that case we do not have an act of virtue,

⁴¹ For more detail, see Annas (1993: chs 1–2) (‘Making Sense of My Life as a Whole’ and ‘The Virtues’).

in the technical sense That is, a virtuous person may successfully achieve heroverall aim, but fail to perform an act of virtue in the technical sense becauseshe fails to achieve her immediate target The person is virtuous, and acts virtuously, from virtuous motivation and reasoning Yet the act fails to be anact of virtue in the technical sense because of facts that have nothing to do withvirtue This is surely a strange and undesirable result Admittedly an act ofvirtue in the technical sense is a term of art; but surely it is strikingly odd that

to succeed in performing one you need precisely what virtue does not supply.This second problem makes it particularly clear that an act of virtue in thetechnical sense contains tensions which arise from the conflicting demandshere of ethics and of epistemology To serve in a definition of knowledge, the act

of virtue in this sense must guarantee success in achieving its immediate target But precisely this makes it utterly unlike acts of virtue within virtueethics, for which the relevant kind of success is success in achieving the overallaim, even when the immediate target is unavoidably missed

IV

I have stressed two aspects of the structure of virtue, as that figures in theancient virtue ethics tradition as a whole and not just in Aristotle I have nothere, of course, had the scope to develop them in a way adequate to makethem appear appealing, still less compelling, though a fuller expositionwould, I think, show that they both have individual advantages and form part

of a powerful and attractive type of moral theory

I have argued that if we take these two aspects of virtue—skill and success—seriously as they figure in the virtue ethics tradition, we find reason to doubtthat virtue as it figures in that tradition can unproblematically be used as abasis for a traditional definition of knowledge Tensions emerge between thestructural requirements of virtue in the virtue ethics tradition and the struc-tural requirements of virtue as it figures in virtue epistemology In itself this isfar from fatal to virtue epistemology, or to the project of defining knowledge

in terms of intellectual virtue It simply shows that virtue epistemology willhave some problems as long as it works with the notion of moral virtue whichcomes from the ancient tradition of virtue ethics

There are many possible responses to this One might be to retain the goal

of using virtue as the basis for a definition of knowledge, but to work with

a notion of moral virtue which is explicitly a product of modern reflection,